In honor of Mathematics and Statistics Awareness Month, SIAM is spotlighting mathematicians and statisticians throughout April. We sat down with Dr. Harvey J. Stein, SIAM member and Head of the Quantitative Risk Analytics Group at Bloomberg, to find out more about how applied mathematics and statistics are used in financial math and how he and his team use math in their work.
Dr. Stein is well known in the industry, having published and lectured on mortgage backed security valuation, credit valuation adjustment (CVA) calculations, interest rate and foreign exchange (FX) modeling, credit risk modeling, financial regulation, COVID-19 data analysis, and other subjects. Dr. Stein is on the board of directors of the International Association for Quantitative Finance (IAQF), a board member of the Rutgers University Mathematical Finance program, an adjunct professor at Columbia University, and organizer of the IAQF/Thalesians financial seminar series. He's also worked as a quant researcher on the Bloomberg for President campaign. He received his B.A. in mathematics from Worcester Polytechnic Institute (WPI) in 1982 and his Ph.D. in mathematics from UC Berkeley in 1991.
Who are you and what do you do? How does your work impact everyday life?
In Quantitative Risk Analytics at Bloomberg, we provide models for analyzing credit risk and market risk.
In the broadest scope, risk management is an essential part of decision-making. Every decision that a person makes requires weighing risks and rewards. As such, risk management is really a fundamental part of everyday life.
Risk management requires formalizing the analysis of these risks. This breaks down into three steps - identifying what can go wrong, assessing the probability and impact of these events, and mitigating the impact. The impact is mitigated by either reducing the probability, reducing the impact, or monitoring the situation so that action can be taken in the future if necessary.
Financial risk management is about applying risk management to financial markets. When sound financial risk management practices are not followed, catastrophic events sometimes occur. One example is JPMorgan losing $6 billion due to trades made by the "London Whale." Another example is Knight Capital almost going bankrupt due to an operational failure. More recently, we've seen $10 billion in losses due to the failing of Archegos Capital Management, and the Robinhood fiasco with Gamestop.
If sound financial risk management practices were followed by businesses, the financial crisis of 2008 might not have occurred.
How do you personally use statistics in your work?
The market risk models are statistical in nature. They require estimating the loss distribution of a portfolio of financial instruments. The value of a portfolio is a function of a large number of factors, so much of the work is about developing good estimates for the joint distribution of these factors.
The credit risk models are highly statistical. The try to estimate the probability of a given company defaulting or going bankrupt. They are all ultimately regression based models. Much effort is expended on classical statistical work — collecting data, cleaning data, identifying appropriate factors, performing regressions, and analyzing results.
More generally speaking, how is statistics used in financial math?
In addition to risk management models, trading strategies are also statistical in nature. Quantitative hedge funds and index tracking mutual funds all use statistical methods to manage their portfolios. And machine learning is essentially statistical in nature, and its usage continues to grow.
Thank you, Dr. Stein, for your contributions to SIAM and for your leadership in mathematics and statistics!